Age-related abnormalities of thalamic shape and dynamic functional connectivity after three hours of sleep restriction

Introduction

Sleep loss is a common and serious issue that has exerted a negative effect on health nowadays. Individuals suffering from sleep loss have a low quality of life and a high likelihood of experiencing accidents. Insufficient sleep has been associated with the impairment of cognition and emotion (Lim & Dinges, 2010). The sleep–wake cycle has been observed to be disrupted in psychiatric and neurological disorders (Videnovic et al., 2014). Thus, elucidating the mechanisms underlying sleep deprivation is an important goal in basic and clinical neuroscience.

Research has suggested that insufficient sleep is involved in the dysfunction of the thalamus. Neuroimaging task studies have reported that the abnormal activity of the thalamus after total sleep deprivation is associated with impaired sustained attention (Chee et al., 2010; Chee et al., 2008). Resting-state functional magnetic resonance imaging (fMRI) technique, which captures intrinsic brain activity, has caused lots of attention (Liu et al., 2013; Liu et al., 2012). For example, studies on resting-state fMRI have observed alterations in the amplitude of low-frequency fluctuations (ALFF) of the thalamus (Chen, Qi & Zheng, 2018) and decreased thalamic functional connectivity (FC) with cortical brain areas, such as the anterior and posterior cingulate cortex (Yeo, Tandi & Chee, 2015) and the superior and medial prefrontal cortex (Shao et al., 2013). Moreover, the FC between the thalamus and the default mode network can predict worse working memory performance following sleep deprivation (Chengyang et al., 2017; Lei et al., 2015). However, these studies are all based on the assumption that the resting-state FC is stationary during scanning, thus ignoring the dynamic features of FC on a smaller time scale (Allen et al., 2014; Preti, Bolton & Ville, 2017). Sleep deprivation studies have shown altered dwell time and transit between FC states using whole brain dynamic FC analysis (Li et al., 2020; Xu et al., 2018). Hence, the effects of sleep deprivation on the dynamics of thalamic FC remain unclear.

Additionally, substantial evidence has suggested that individuals with different age cohorts have different sensitivities to losing sleep. For example, old people tolerate sleep deprivation better than young people (Duffy et al., 2009). The decrements of bad cognitive performance after insufficient sleep are higher in young adults than in old adults (Richards et al., 2017). A verbal encoding task reported an increased activation of the anterior parahippocampus following sleep deprivation in old adults; however, this activation decreases in young adults (Jonelis et al., 2012). Studies on resting-state fMRI also have reported altered FC of insula after sleep deprivation (Long & Cheng, 2019) and the association of FC of the medial temporal lobe with sleep quality only in young people (Liu et al., 2018). Nevertheless, research has not clearly determined whether age could affect the alteration of thalamic functional dynamics caused by sleep deprivation.

The brain structure has been thought to constrain the brain function significantly (Honey et al., 2009). Thus, the structural substrate of brain dysfunction should be studied to gain full understanding of the mechanism underlying sleep deprivation. Structural MRI studies have already observed gray matter volume loss in the thalamus following sleep deprivation by using voxel-based morphometry (VBM) (Liu et al., 2014; Long, Cheng & Lei, 2020). Currently, a vertex-based shape analysis has emerged to investigate the local atrophy in subcortical structures (Patenaude et al., 2011). This method overcomes some limitations of VBM, such as reliance on the extent of arbitrary smoothing (Patenaude et al., 2011). Thus, researchers believe that local subcortical abnormalities are detected more precisely in the shape analysis than in VBM (Kim et al., 2013). Indeed, some studies have demonstrated subcortical structural abnormalities using shape analysis, which were not detected in VBM (Kim et al., 2013; Štěpán Buksakowska et al., 2014). Shape analysis has revealed alteration of subcortical regions in sleep disorders (Macey et al., 2018; Rahayel et al., 2018). However, whether thalamic shape is altered following sleep deprivation remains unclear.

Previous studies that focused on partial sleep deprivation detected deprived alteration of intrinsic FC and activation of thalamus (Nilsonne et al., 2017; Tamm et al., 2017). However, research has not comprehensively explored whether the thalamic shape and dynamics are altered after partial sleep deprivation. By combining vertex-based shape analysis and resting-state dynamic FC analysis, the current study investigates age-related local structural abnormalities of the thalamus and its dynamic FC changes after three hours of sleep restriction. On the basis of previous findings, we hypothesize that thalamic atrophy and age-related alteration of FC variability are observed after three hours of sleep restriction.

Materials and Methods

Results

The subject screening procedure can be seen in Fig. 1. Five subjects were excluded because they did not undergo T1 image scanning before or after sleep restriction. One subject was excluded due to artifacts. Finally, 55 participants remained (23 old adults, 32 young adults).

The sleep diaries revealed that young and old participants had normal sleep times of 8.46 ± 0.79 h and 8.34 ± 0.85 h, respectively. Ambulatory polysomnography data showed that the total sleep times in the sleep restriction condition were 2.97 ± 0.59 h for young adults and 2.52 ± 0.4 h for old adults. In the normal sleep condition, the total sleep times were 6.71 ± 1.48 h and 5.56 ± 1.34 h for young and old groups, respectively. The PANAS score was recorded in both conditions. So, mixed ANOVA (two factors: age and deprivation) was used to test difference in the PANAS scores. For a positive PANAS score, we only observed the significant main effect of deprivation (F = 5.36, p = 0.025) with a decreased PANAS positive score after sleep restriction. We did not find any significant effects for the PANAS negative score. No significant differences (p > 0.05) were observed in the gender, ESS, KSQ sleep quality index, and KSQ snoring symptom index between young adults and old adults. The old adults had a significantly (p < 0.05) lower ISI score than young adults (Table 1).

Shape analysis revealed the significant main effect of deprivation with decreased projection value in the left thalamus after three hours of sleep restriction (Fig. 2). No significant main effect of deprivation was observed in the right thalamus, and no significant age × deprivation interaction effect was present in the bilateral thalamus. The changes of projection value were not correlated with participant measures.

By using resting-state temporal dynamic FC analysis, we observed a significant age × deprivation interaction effect of the left thalamic FC variability. Post-hoc analysis showed that three hours of sleep restriction decreased the FC variability between the left thalamus and the left superior parietal cortex (SPC), which was observed only in young adults (Fig. 3, Table 2). The reduced variability of the left thalamic FC with the left SPC was still found when the window lengths of 20 TR and 40 TR were used (Figs. S1 and S2). We did not find significant main effect of deprivation. Additionally, a significantly negative linear correlation (r = −0.34, p = 0.01) was observed between the ISI score and the changes in variability (post-deprivation minus pre-deprivation) of the left thalamic FC with the left SPC (Fig. 4).

Discussion

The current study investigated age-related local structural abnormalities of the thalamus and the changes of the thalamic FC variability after three hours of sleep restriction. We found significant atrophy in the left thalamus and reduced FC variability between the left thalamus and the left SPC in young adults. Additionally, the change of FC variability between the left thalamus and the left SPC was negatively correlated with the ISI score.

The thalamus represents a pivotal gating hub, which sends arousal signals from brainstem to cortical brain areas. Functional neuroimaging studies showed that impaired sustained attention after sleep deprivation was involved in abnormal thalamic activity. Such abnormal activity of the thalamus resulted in altered ascending arousal signals and then affected the cortical attentional networks (Doran, Dongen & Dinges, 2001; Goel et al., 2009). However, the direction of the activity changes within the thalamus was not consistent. Some studies have reported greater sleep-deprived activity (Chee & Choo, 2004; Chee et al., 2010), whereas other studies have found diminished thalamic activity (Chee & Tan, 2010; Chee et al., 2008). These inconsistences may be explained in the context of different task performances (Chee & Tan, 2010; Goel et al., 2009; Thomas et al., 2000). Structural neuroimaging studies have further reported reduced gray matter volume of the thalamus after sleep deprivation (Liu et al., 2014; Long, Cheng & Lei, 2020), thus providing structural evidence for thalamic dysfunction. The thalamic atrophy observed in the current study is in line with previous findings, thus strongly suggesting sleep-deprived dysfunction of the thalamo-cortical connectivity.

In addition, we observed a significant age × deprivation interaction effect with reduced FC variability between the left thalamus and the left SPC after sleep restriction in young adults. A previous study did not report altered thalamo-parietal FC after sleep deprivation using stationary FC analysis (Shao et al., 2013). However, the reduction of the thalamo-parietal FC variability observed in the current study might provide new insight into understanding the mechanism of sleep deprivation. The result was consistent with a study reporting a decreased number of transitions between brain states, which suggested reduced cognitive flexibility after sleep deprivation (Teng et al., 2019). The SPC is one of key regions of the fronto-parietal network (FPN), which has the ability to adapt to a wide variety of tasks dynamically and flexibly. Damages to the FPN may lead to cognitive control dysfunction, such as impaired attention and working memory, which was observed in sleep-deprived individuals (Chee et al., 2010; Choo et al., 2005). The decreased variability of the thalamo-SPC FC following sleep restriction found in the current study might indicate reduced ability to switch between tasks. Moreover, we observed a negative correlation between the changes in the thalamo-SPC FC variability and the ISI score, thus indicating that more severe insomnia entails more reduction in the variability of thalamo-SPC FC following sleep deprivation. This finding suggested that insomnia might be associated with dysfunction of the SPC.

Interestingly, the sleep-deprived reduction of FC variability was not found in old adults. A behavioral study found that old adults had less sleep requirement during daytime and reduced nocturnal slow-wave sleep compared with young adults (Dijk et al., 2010). Thus, young people were more sensitive to sleep deprivation than old people (Duffy et al., 2009). In addition, Zitting and colleagues (Zitting et al., 2018) found that young adults were more vulnerable to chronic sleep deficiency and recurrent circadian disruption and especially had more attentional failures than older adults. The higher ISI score in young adults observed in the current study provided direct evidence of this supposition. These findings were in accordance with previous neuroimaging studies, thus suggesting that structural and functional abnormalities resulting from sleep deprivation were age-dependent (Long & Cheng, 2019; Long, Cheng & Lei, 2020; Richards et al., 2017; Rogasch et al., 2009).

Several limitations in current study need to be mentioned. First, the present study might have limited statistical power due to small sample size. Future researches should involve a larger sample size of participants. Second, the KSQ sleep quality score and ISI score were not recorded after sleep restriction. Thus, we can not investigate the relationship between sleep deprived thalamic alteration and changes of sleep quality. Future studies are warranted to include the sleep quality index in both conditions.

Conclusions

In conclusion, the current study investigated age-related changes in the thalamic shape and FC variability after three hours of sleep restriction. We found significant local atrophy in the thalamus, and an age × deprivation interaction effect with reduced variability of FC between the left thalamus and the left SPC following sleep restriction only in young adults. These results highlighted the role of the effect of age in studies on sleep deprivation.

Supplemental Information